PREDICTIVE ANALYSIS IN PREVENTING MISHAPS
-Summary of the article based on the transcript of Dr Data Show episode, “Five Ways Your Safety Depends on Machine Learning “
Predictive analysis is a technique of using existing data to predict what could happen in future.
One application of this technique is in using predictive analysis as a norm in safety standards by guiding concerned officials to flag risky bridges, buildings for inspection, thus preventing mishaps and thereby saving lives.
“A computer program is said to learn from experience E with respect to some class of tasks T and performance measure P if its performance at tasks in T, as measured by P, improves with experience E.” – TOM M MITCHELL. A widely used quote on machine learning.
Historic data can help to identify deteriorating bridges and risky based on safety parameters. With predictive analysis, officials can inspect these infrastructures before any mishaps happen and hence lower the risk of accidents. As mentioned in the article, “Fortify buildings, bridges, and other infrastructure” as a safety step by identifying the risk and prioritizing the inspections based on risk levels calculated. Machine learning can be implemented using images of bridges to detect cracks and hence help in shortlisting the structures likely to have cracks in the future. It can also be used to identify manholes with higher risk for incidents like fire or explosions. Moreover, City of Chicago officials have used predictive analysis to identify households with greater risk of lead poisoning and then flagging the area before poisoning happens.
Safety is priority. We all want a safe world to live in. As suggested in the article, we can use predictive analysis in addition to other risk management methods to ensure this. Nothing is 100% guaranteed, however a step towards good standards is always a better choice.
-By Varsha Syam Sundar